Use cohorts to see whether your store creates lasting customers

You have a good month. Sales rise, your launch email performs well and the dashboard looks encouraging. The following month is quieter, so you start wondering whether the first result was a useful signal or just a temporary spike. A cohort view gives you a more careful way to answer that question: compare buyers who first purchased in the same period, then follow what those buyers do afterwards.

What a customer cohort shows

A cohort is simply a group defined by a shared event. For a digital storefront, the most practical event is often a customer’s first purchase. Your January cohort contains everyone whose first order arrived in January. Your February cohort contains everyone whose first order arrived in February. A person belongs to one first-purchase cohort, even if they buy several times later.

Build a useful cohort view

Start with a spreadsheet or the customer records available in your storefront. You need a customer identifier, first-purchase date, order dates, order values and the product in each order. If you sell through digital downloads, a customer’s email can usually act as the identifier, provided you handle aliases and duplicate records consistently.

1. 1Choose the cohort event. First purchase by calendar month is a sensible starting point. For a short launch, use launch week or launch date instead.
2. 2Assign each buyer to the period of their first purchase. Do not move them into a new cohort when they buy again.
3. 3Choose follow-up windows, such as 30, 60 and 90 days after the first order. Keep the windows consistent across cohorts.
4. 4Count the original buyers in each cohort. This is your denominator and prevents a group of five buyers being compared directly with a group of fifty.
5. 5Count how many people in each cohort made another purchase within each window.
6. 6Add the later revenue from those cohort members, and record which product or offer produced the repeat order.

The basic repeat-purchase calculation is straightforward: repeat buyers in the window divided by original buyers in the cohort, multiplied by 100. If 12 of 40 January buyers purchase again within 60 days, the 60-day repeat rate is 12 ÷ 40 × 100 = 30%. Keep the number of original buyers beside the percentage, because percentages from very small cohorts can look more certain than they are.

Read the pattern, not just the percentage

Look across cohorts and down the follow-up windows. Across a row, you can see whether one customer group keeps returning. Down a column, you can compare the same age of different cohorts: for example, each group’s 30-day repeat rate. This is more useful than comparing January’s total revenue with February’s total revenue, because each month may contain a different mix of new and returning buyers.

A cohort may have a strong first month and weak later activity. That can mean the launch message was effective at creating an initial purchase, while the product has no natural next step. The reverse can also happen: a modest first sale can lead to several later purchases if the first product points clearly towards a related template, course or membership. If you offer memberships, separate renewal behaviour from one-off product purchases so the mechanism is not hidden.

Turn the finding into a store decision

• If one cohort returns more often, inspect what was different about its entry product, price, promise, traffic source or follow-up sequence.
• If buyers return but choose a particular second product, make that next step visible on the first product page and in the post-purchase message.
• If a launch cohort buys once and disappears, ask whether the offer was deliberately designed as a one-off. Not every product needs to create repeat purchases.
• If later revenue comes mostly from discounts, compare the value of the repeat order with the margin you gave away. A higher repeat rate is not automatically a better result.
• If a cohort is too small to interpret, combine several comparable periods or wait for more buyers rather than making a precise claim from a handful of orders.

Use cohorts without creating a second job

You do not need to monitor a cohort table every day. Choose one review point after each month or launch. Record the cohort size, repeat buyers, repeat rate, later revenue and the products bought next. Then write one sentence about what you will change or test. A simple routine keeps the analysis connected to a decision rather than turning it into another dashboard to maintain.

If you sell from store.fan’s built-in analytics, use the available sales information as the starting point for this review, then export or organise the customer-level detail you need for your chosen windows. Store.fan keeps payments connected to your own Stripe account, so Stripe processing fees still apply, but store.fan takes 0% of sales. The practical cost is your time, plus any paid plan you choose after its 14-day trial.